His interest in the concept of surprise and how to avoid it followed a circuitous path. As a professor at Harvard University's Kennedy School of Business, Watkins had lectured on international diplomacy, which evolved into an interest in corporate diplomacy—that is, how senior executives shape their environment. That, in turn, evolved into an interest in diplomacy amid crisis and crisis management. After that, he remembers, "I started teaching a couple of class sessions about surprise as a phenomenon, and why people are surprised."

In this Q&A with consulting Web editor Howard Baldwin, Watkins talks about the genesis of his book, Predictable Surprises: The Disasters You Should Have Seen Coming, and How to Prevent Them. Co-authored with Harvard Business School professor Max Bazerman, it was published in October by the Harvard Business School Press.

Q: Beyond your class lectures, was there something specific that inspired your book?

A: I came upon a book written in 1962 about the attack on Pearl Harbor, Warning and Decision by Roberta Wohlstetter. It was a brilliant account of why we were surprised, despite the availability of information about an attack. We had warnings, and we were reading the Japanese communications, but we made some bad assumptions.

The United States didn't think the Japanese could attack us in Pearl Harbor with torpedoes, because when a torpedo is launched, it goes down in the water, then comes up and runs for a period before it arms. But that assumption proved erroneous, because the Japanese were feverishly figuring out how to build torpedoes that stayed shallow and armed fast. The book also cites organizational reasons why we were surprised: The left hand didn't know what the right hand was doing.

Q: Your co-author, Max Bazerman, was also thinking about the topic, but in a different way.

A: Max testified before the Securities and Exchange Commission in 2000, warning about auditor independence, long before Enron, MCI, and Tyco proved him right. He's a cognitive psychologist, a specialist in decision making. He believed the system was a train wreck waiting to happen. He saw the signs, so why didn't others? Then 9/11 happened, and we started doing more thinking on the subject.

Q: Was 9/11 just a failure of knowledge management?

A: There's a big component of that. Integration [lapses] are a classic aspect of knowledge-management failure. Different parts of the organization know different things, but there's no integration. Before 9/11, the National Security Agency, the FBI field offices, and the CIA all knew certain facts, but there was no mechanism for pulling them together. There had been some, but they'd been undermined by budget cuts. There was no institutional memory. For knowledge management to work, you have to create institutional memory or avoid memory loss. Because of the way people were being shuttled through positions in the federal government, there was no institutional memory.

Q: What's the take-away for a business leader, and especially a CIO?

A: There are lots of business cases in the book. Take the blackout of 2003. That was predictable because of underinvestment in traditional infrastructure and who controlled the electrical lines. Some people had figured out that there were risks with the grid, and some hadn't. That situation goes directly to the heart of IT. Some people have backup data centers, and some just have uninterruptible power systems. The New York Stock Exchange and Nasdaq shut themselves down and opened up the next morning without a hitch. But the American Stock Exchange was stuck. Same with the airlines—U.S. air traffic went without a hitch, and Canada shut down.

In the book, we differentiate between companies and managers who are aware of and act to prevent predictable surprises and those who don't. We try to lay out a standard of responsible leadership. The core responsibility is to scan external and internal environments to figure out where threats are coming from, which are the most important, and how to tackle them. The mantra in the book is recognize, prioritize, and mobilize.

Q: But you can't hold someone responsible for something that's truly a surprise.

A: We set a high bar. You're right—you don't hold people responsible for an unpredictable surprise. But when people don't put capital behind something that's foreseeable, that's negligence. There are ways to set up systems, and for IT, there are ways of intelligence gathering and building coalitions, both internally and beyond your organization.

Q: In a way, this sounds like the discussion about how much insurance to buy.

A: That's right, and people argue against the cost of insurance. But what do you do when a change will have a broad, significant impact, but with an immediate cost? That's why too often it takes things going to crisis and blowing up before something is done. That's why you have to be strategic in your coalition building.

Q: Getting back to knowledge management, will we ever be good at it?

A: In the mid-1990s, I was doing thesis research on knowledge management at Ford Motor Co. I was trying to design systems that would capture information about problems and their impact, so engineers could consult the system and avoid problems in the future. They were cycling young engineers through quickly, and they were functionally organized, rather than team-based. We built wonderful systems where people could input information, but they didn't work, because no one would input information.

Q: Did anything good come out of your work at Ford?

A: Actually, yes. In the process, we created a new, team-based organization, which did work. The organization became more important than the tool. The technology wasn't quite there in the mid-'90s. Now that they have a good team structure, it would be interesting to go back today. We were a little ahead of the curve in terms of what was commercially possible.

Q: What lesson do you take from that?

A: In order to work, knowledge-management systems have to be kept live, and there have to be incentives to keep them live. In some sense, the most important source of knowledge is inside people's heads. You don't want an organization that's leakier than necessary. You don't want too many people cycling out. You need to set up team structures, so that the more experienced people teach the less experienced ones. This is going to become a bigger issue as we come up against large numbers of people retiring.

Q: How can IT help with this?

A: Simplicity is key. IT can help create short report forms that go into a database. The technology is really only an enabler. You can leverage the capacity of IT, which is memory—IT doesn't forget. Then you work through the access problems. But if you don't marry organizational structure, disciplines, incentives, and systems together, you'll never be successful.

Q: Do you think people ever learn from their mistakes?

A: Yes. Whether organizations learn from their mistakes is a different question. There are so many ways things can go wrong. Organizations can fail to learn or, having learned things, they can forget them. No one has the incentive to push the process, or everyone thinks it's someone else's responsibility to do it, so you don't capture insights. Organizations have to learn from their mistakes because doing so is the beginning of wisdom.